To project overview

New possibilities in yeast breeding: robots, AI and wild yeast

Annual Review Article 2020

In 1875, J.C. Jacobsen set up the highly visionary Carlsberg Laboratory (now the Carlsberg Research Laboratory), which became a department of the Carlsberg Foundation following the latter’s establishment in 1876. Today, 144 years later, Carlsberg is clearly distinguished from other major global brewery groups by its dedication to beer research and innovation. The Carlsberg Research Laboratory is unique because its research areas cover the entire beer brewing process, from brewery technologies, product quality, sustainability, raw materials and ingredients to yeast and alternative brewing organisms.

By Claes Gjermansen, senior researcher, Jochen Förster, professor and director, Birgitte Skadhauge, vice president, adjunct, professor Carlsberg Research Laboratory

Carlsberg is also a pioneer in the brewing industry when it comes to yeast, barley and hop genome sequencing. 

This insight and knowledge, and especially the detailed understanding of genomes and genes in brewing ingredients, will ensure that Carlsberg is also able to continue producing new, groundbreaking innovation in the future. Furthermore, the Carlsberg Research Laboratory carries out basic research that will support Carlsberg’s general business strategies going forward. 

This article gives an insight into how Carlsberg’s yeast research is driven by deep research insight complemented by more than 170 years of brewing knowledge.

Looking back, there are numerous examples of industrial companies emerging, becoming large and important, and then, following a period of greatness, stagnating and being overtaken by other more dynamic and expansive competitors. 

Carlsberg is also a pioneer in the brewing industry when it comes to yeast, barley and hop genome sequencing

Because staying at the top requires adaptability, for which mastering the latest technology and knowledge is essential. And contributing to this is precisely the purpose of the Carlsberg Research Laboratory. J.C. Jacobsen saw from the very start of Carlsberg that basic knowledge is the platform for technology. 

He therefore sought to link up with the leading scientists of the time in Denmark and abroad. And he was able to understand their way of thinking. 

When the Carlsberg Laboratory was founded, he appointed Johan Kjeldahl and Emil Christian Hansen as professors to study malting processes and fermentation processes respectively. Both scientists became world famous.

Wild Saccharomyces tamed at Carlsberg

The main ingredients of today’s beer, besides water, are the same as when Carlsberg was founded: barley, hops and yeast. So it is still the same organisms that are the subject of research, albeit the means and methods have evolved and most things can be done better and faster than a hundred years ago. 

Yeast, however, is not just yeast. There are many different varieties. 

Emil Christian Hansen revolutionised microbiology and the brewing industry by isolating the world’s first pureculture yeast, Unterhefe No. 1, which hugely improved the quality of beer and, consequently, its popularity. 

The Carlsberg Research Laboratory’s next major contribution to the understanding of the biology of brewing came with Øjvind Winge, an exceptional and versatile geneticist. 

Winge was appointed as a professor at the Carlsberg Laboratory in 1933 and held the post until 1956. He contributed groundbreaking work in the understanding of the chromosome number of plants, hop genetics and, in particular, yeast genetics. 

He demonstrated that Saccharomyces yeasts are able to reproduce sexually and that the genetics follows Mendel’s laws. Sexual reproduction is a necessary precondition for crossbreeding. The discovery was the basis for Saccharomyces cerevisiae becoming one of the most important genetic model organisms internationally. 

It is a requirement of model organisms that they are easy to handle in the laboratory and can live on well-defined media. Also, they should have relatively simple genetics and a short generation time. Saccharomyces cerevisiae is a haplodiplont. 

This means that the vegetative cells contain two sets of chromosomes. During meiosis in sexual reproduction, sex cells are formed that contain half the chromosome number, i.e. one set of chromosomes. There are 16 chromosomes in a chromosome set in Saccharomyces cerevisiae. 

In 1992, chromosome III became the first completely DNA-sequenced chromosome. The Carlsberg Laboratory made a huge contribution to this, and in 1996 the entire genome was sequenced. Saccharomyces cerevisiae was thus the first eukaryote, i.e. an organism whose cells have a nucleus, in which this was done. 

Unlike most other eukaryotes, Saccharomyces has a compact genome. Most of the DNA are coding sequences. 

Only 280 out of approximately 6,000 genes contain so-called introns, which are non-coding sequences that the cells splice out during transcription, i.e. translation between gene and protein sequence. This is in contrast to barley, where most of the DNA is non-coding. 

STAY CURIOUS: Crops of the future

The key importance of Saccharomyces cerevisiae in genetic and biochemical basic research is not specifically due to its relevance to beer. It is due to the fact that yeast research can make us more knowledgeable about functions in other organisms, e.g. humans, because yeast and humans have many essential cell functions in common. 

Medical research has been helped by experiments with yeast, which allow experiments that would be ethically impermissible with humans. Carlsberg’s research was significantly boosted with the founding of the Carlsberg Research Centre in 1976. 

A yeast research group was established, and applied yeast breeding projects were started up. Lager yeast was previously regarded as being sterile, rendering crossbreeding impossible. 

The researchers discovered that growth on a starvation medium at lower temperatures led to low-frequency spore formation in Carlsberg’s yeast. A small proportion of these spores were capable of germinating. 

Some of the spore clones that were generated had a stable pairing type. This meant that, for the first time, crossbreeding was possible. 

In connection with this, it was discovered that Carlsberg’s production yeast was genetically more complex than laboratory yeast, comprising a genetically complex allotetraploid species hybrid to which Saccharomyces cerevisiae has contributed half the chromosomes. The other half originates from a species of yeast unknown at the time, Saccharomyces eubayanus.


As well as making alcohol, yeast contributes to the taste of beer. Different strains produce different aroma components. These can be fruity aromas, which in many cases are characteristic of a particular product brand. Not all aromas from yeast are regarded positively; ”off-flavours” are occasionally formed under particular conditions. 

In most cases, the biochemistry, synthesis pathways and genetics behind these components are known. What may be an off-flavour in one beer type may be a requirement in another type. One example of this is 4-vinylguaiacol. The flavour, which is vanilla- or clove-like, is characteristic of wheat beer. It is not tolerated in pilsner beer. 

Each time Carlsberg has acquired a new brewery, yeast strains have been included in the acquisition. Carlsberg therefore has an extensive collection of strains.

Different yeast strains have different technological properties and contribute different aroma components. Analogously to barley breeding, properties from one yeast can often be combined with properties from another yeast in crossbreeding. 

When a new project starts up, the questions are: which product, which form of production, which raw materials and which flavours will the yeast be used for? Suitable breeding strains are then selected based on descriptions of strains in the collection and new laboratory screenings. Crossbreeding produces a large number of crossbred products. 

The laboratory has robots that can pick up yeast colonies. These can be automatically transferred to selection plates, from where they can be transferred to microtiter plates containing different liquid media. 

The growth in these can be tested in a Growth Profiler (fig. 1.). This provides data on which sugars can be fermented and at what rate. After this first step, the unsuitable isolates can be rejected, while any that are judged to be promising are tested in larger volumes. The next steps will be fermentations in Carlsberg’s pilot brewery in 50- or 200-litre tanks. 

Fig. 1 (Left to right) Yeast cells are incubated in EnzyScreen plates in microtiter wells filled with wort. The plates have a transparent bottom, and this allows taking pictures of cell densities over time. Using the Growth Profiler, the digital pictures are translated into G-values, which is a measure for biomass. G-values over time are visualized in curves and are compared with a reference yeast. The slope can be used to estimate yeast growth. When the green value flattens, it can be used to estimate the degree of fermentation.
Photo: Ross Fennessy

The beer can then be analysed and judged by a tasting panel. The final evaluations will be conducted at full scale in one or more breweries. A decision can then be taken as to whether the strain is suitable for production. 

All yeast breeding at Carlsberg is non-GMO, and this has been successful. Today, all Carlsberg Pilsner is brewed with a strain that was developed in the Carlsberg Research Laboratory. It was created with the aid of mutagenesis, selection and crossbreeding. 

The strain develops less diacetyl, an unwanted taste component. This means that the beer’s storage time can be shortened and taste faults minimised. A similar project with an improved Tuborg yeast has also been successful. A new yeast strain for Tuborg Pilsner is currently being phased in at our breweries.

Wild Brettanomyces tamed at Carlsberg

In 1894, Emil Christian Hansen employed Niels Hjelte Claussen as an assistant at the Carlsberg Laboratory to work with the biochemistry and microbiology of beer. 

Four years later, he became the director of New Carlsberg’s operating laboratory, where he continued his microbiological work. From English porter, he isolated and cultivated a new ”wild” yeast species, which he called Brettanomyces (British fungus). It was patented in 1904. 

Through controlled use of this species, he was able to produce porter with a typical English stamp. Brettanomyces is widespread in the wine and beer industry, regarded either as an impurity or a contributor of positive aromas. 

It is not just the forms of production that are changing; so too are the products. There is an ever-increasing focus on specialty products. 

In this regard, Brettanomyces is gaining in popularity. In one project, the Carlsberg Research Laboratory has genome-sequenced a large collection of Brettanomyces strains from all over the world and analysed their production of aroma components. 

In this way, strains have been identified that will be of interest for the production of new specialty products. 

 Fig. 2 Shows a cladogram of 84 Brettanomyces strains, correlating genetic organisation with relevant aroma components.

Fig. 2 Circular cladogram of 84 Brettanomyces strains. The phylogentic tree was produced by comparison of the predicted proteins in the whole-genome of each strain. The origin of the strains, among others, are indicated in different colours.
Adopted from: Colomer MS, Chailyan A, Fennessy RT, Olsson KF, Johnsen L, Solodovnikova N and Forster J (2020) Assessing Population Diversity of Brettanomyces Yeast Species and Identification of Strains for Brewing Applications. Front. Microbiol. 11:637. doi: 10.3389/fmicb.2020.00637.

Robots, artificial intelligence and yeast breeding

Current technology allows the production of an almost unlimited number of new crossbreeds. The selection is carried out by researchers based on analysis results. 

The achievement of a sufficient number of analysis results is probably the narrowest bottleneck in the laboratory’s work. The chemical analyses require a comprehensive effort with advanced measuring equipment, and not everything can be analysed. 

The content of aroma components in beer is extremely complex. The human sense of taste is based not only on the quantity of the various components, but to a large extent also on the relationship between them. 

Some aromas can dampen or intensify the perception of others. There are as yet no quick methods that can give a complete overview. A quantum leap forward could be the Beer Fingerprinting Project, the idea for which came from Jochen Förster, Professor and Director, Carlsberg Research Laboratory. 

The project is a collaboration with the Interdisciplinary Nanoscience Center (iNANO) at Aarhus University (development of advanced sensors), DTU Chemical Engineering (implementation of different fermentation scenarios) and Microsoft (signal analysis, including machine learning algorithms). The project is supported by Innovation Fund Denmark. 

The objective is to scale up yeast breeding, implement new products and reduce the time from the laboratory bench to production and sale. It is likely that the project, in addition to achieving this objective, will bring some new, unexpected observations that can be applied to novel projects. 

The current status is that sensors have been developed that are able to distinguish between Carlsberg Pilsner, Tuborg Pilsner, Wiibroe Pilsner and Carlsberg Nordic.